Association between urinary sodium and circulating lipid levels: a Mendelian randomization study.

Abstract

Urinary sodium was indicated to be associated with dyslipidemia, but inconsistent conclusions for this association exist across the present observational studies. This study aimed to evaluate the causal association between urinary sodium and circulating lipid levels [low-density lipoprotein cholesterol (LDL-C), triglycerides, and high-density lipoprotein cholesterol (HDL-C)] through Mendelian randomization. Univariable Mendelian randomization (UVMR) and multivariable Mendelian randomization (MVMR) with pleiotropy-resistant methods were performed. Data for urinary sodium were obtained from the genome-wide association study (GWAS) from 446,237 European individuals. Data for lipid profiles were extracted from GWAS based on the UK Biobank (for the discovery analysis) and the Global Lipids Genetics Consortium (for the replication analysis). In the discovery analysis, UVMR provided evidence that per 1-unit log-transformed genetically increased urinary sodium was associated with a lower level of HDL-C level (beta = -0.32; 95% CI: -0.43, -0.20; p = 7.25E-08), but not with LDL-C and triglycerides. This effect was still significant in the further MVMR when considering the effect of BMI or the other two lipid contents. In contrast, higher genetically predicted triglycerides could increase urinary sodium in both UVMR (beta = 0.030; 95% CI: 0.020, -0.039; p = 2.12E-10) and MVMR analyses (beta = 0.029; 95% CI: 0.019, 0.037; p = 8.13E-10). Similar results between triglycerides and urinary sodium were found in the replication analysis. Increased urinary sodium may have weak causal effects on decreased circulating HDL-C levels. Furthermore, genetically higher triglyceride levels may have independent causal effects on increased urinary sodium excretion.